- Email: [email protected]
Latitudinal and seasonal capacity of the surface oceans as a reservoir of polychlorinated biphenyls
Environmental Pollution 128 (2004) 149–162 www.elsevier.com/locate/envpol
Latitudinal and seasonal capacity of the surface oceans as a reservoir of polychlorinated biphenyls Elena Juradoa, Rainer Lohmannb, Sandra Meijera,b, Kevin C. Jonesb, Jordi Dachsa,* a
Department of Environmental Chemistry, IIQAB-CSIC, Jordi Girona 18-26, Barcelona 08034, Catalunya, Spain b Environmental Science Department, Lancaster University, Lancaster LA1 4YQ, UK Received 14 July 2003; accepted 11 August 2003
‘‘Capsule’’: Model calculations estimate the latitudinal and seasonal storage capacity of the surface oceans for PCBs. Abstract The oceans play an important role as a global reservoir and ultimate sink of persistent organic pollutants (POPs) such as polychlorinated biphenyls congeners (PCBs). However, the physical and biogeochemical variables that affect the oceanic capacity to retain PCBs show an important spatial and temporal variability which have not been studied in detail, so far. The objective of this paper is to assess the seasonal and spatial variability of the ocean’s maximum capacity to act as a reservoir of atmospherically transported and deposited PCBs. A level I fugacity model is used which incorporates the environmental variables of temperature, phytoplankton biomass, and mixed layer depth, as determined from remote sensing and from climatological datasets. It is shown that temperature, phytoplankton biomass and mixed layer depth influence the potential PCB reservoir of the oceans, being phytoplankton biomass specially important in the oceanic productive regions. The ocean’s maximum capacities to hold PCBs are estimated. They are compared to a budget of PCBs in the surface oceans derived using a level III model that assumes steady state and which incorporates water column settling fluxes as a loss process. Results suggest that settling fluxes will keep the surface oceanic reservoir of PCBs well below its maximum capacity, especially for the more hydrophobic compounds. The strong seasonal and latitudinal variability of the surface ocean’s storage capacity needs further research, because it plays an important role in the global biogeochemical cycles controlling the ultimate sink of PCBs. Because this modeling exercise incorporates variations in downward fluxes driven by phytoplankton and the extent of the water column mixing, it predicts more complex latitudinal variations in PCBs concentrations than those previously suggested. # 2003 Elsevier Ltd. All rights reserved. Keywords: POP; PCB; Global dynamics; Marine pollution; Fugacity model
1. Introduction The oceans play an important role in controlling the environmental transport, fate and sinks of persistent organic pollutants (POPs) at regional and global scales (Wania and Mackay, 1996; Dachs et al., 2002). Even though POP concentrations in the open ocean are lower than those observed in coastal areas (Iwata et al., 1993; Dachs et al., 1997a; Schulz-Bull et al., 1998), the large oceanic volumes imply that they may represent an important inventory of POPs. This has been confirmed in budgets performed for some marine regions, such as the Western Mediterranean, and may
* Corresponding author. Tel.: +34-93-4006118. E-mail address: [email protected] (J. Dachs). 0269-7491/$ - see front matter # 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.envpol.2003.08.039
be true for the global oceans (Dachs et al., 1997a; Tolosa et al., 1997). Wania and Mackay (1996) suggested that temperature plays a significant role in determining the transport and sinking of POPs in the oceans at the global scale, through the processes of cold condensation, global distillation and latitudinal fractionation. This would result in an inverted concentration gradient of POPs, with low concentrations in the tropics and high concentrations in the polar regions. On the other hand, less volatile compounds are deposited close to their sources, while more volatile compounds travel further toward the poles, where they deposit preferentially due to lower temperatures. Therefore it has been suggested that temperature could largely explain the oceanic concentrations of POPs due to its influence on air–water exchange and other partitioning processes (Wania and Mackay, 1996).
150
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
However, due to the hydrophobic character of POPs, the spatial and temporal variability in the distribution of water column organic matter has an important influence on the oceanic inventory of POPs. Recently, it has been shown that phytoplankton uptake and settling of particulate matter is a driver of the oceanic sink of POPs such as polychlorinated biphenyls (PCBs) and dibenzo-p-dioxins and-furans (PCDD/Fs) (Dachs et al., 1999, 2000, 2002; Skei et al., 2000; Wania and Daly, 2002) and that the deep ocean is an important global sink for the more hydrophobic POPs (Wania and Daly, 2002). Besides the role that organic matter plays on the settling of POPs, no assessment exists of the potential influence of the spatial variability of organic matter and other biogeophysical variables on the capacity of the surface oceans as a reservoir of POPs. The physical and biogeochemical variables with potential influence on the oceanic inventory of POPs, such as phytoplankton biomass and mixed layer depth, show an important spatial and seasonal variability which presumably affects the cycling and long term fate of POPs. Soil and vegetation organic matter, such as forested surfaces, has traditionally been considered in models that study the global distribution of POPs in terrestrial environments (Eisenberg et al., 1998; Scheringer et al., 2000; Wania and McLachlan, 2001). These studies have shown that terrestrial organic matter has an important influence on the dynamics and inventory of POPs. Therefore, it seems logical that high productivity oceanic areas may exert an important influence on the global distribution of POPs. Specifically, this work will be focused in POPs, such as PCBs, that are found preferentially in the gas-phase (Van Drooge et al., 2002). The objectives of this work are to study the influence of environmental variables such as temperature, biomass and mixed layer depth on the reservoir capacity of PCBs by the suface oceans. The role of settling fluxes driven by enhanced phytoplankton productivity is assessed as a key factor affecting the global oceanic inventory of PCBs. The results are compared with those predicted by the global distillation theory under equilibrium conditions. Finally, the predicted inventory of PCBs in the surface ocean is compared with that reported in soils.
2. Data sources and field sampling 2.1. Physical and biogeochemical data The global distribution of sea surface temperature (SST), chlorophyll a concentrations, and the mixed layer depth (MLD) were assessed as potential variables affecting the reservoir capacity of oceans on a global scale. Data on their global distribution and variability were obtained by remote sensing from satellites and climatological datasets. SSTs were obtained from the Along Track Scanning Radiometer (ATSR) installed in
the European Space Agency ERS-2 satellite (ATSR project web page http://www.atsr.rl.ac.uk/). SST images consist of monthly averaged data with a resolution of a half degree and an accuracy of +/0.3 K. Chlorophyll a concentrations were estimated from fluorescence signals obtained from the Sea-viewing Wide Field-of-view Sensor (SeaWIFS) mission funded by NASA’s Mission To Planet Earth (MTPE) Program (http://seawifs.gsfc.nasa.gov/SEAWIFS.html). Data are averaged monthly at 1 1 resolution and an estimated error of 15%. Chlorophyll a concentrations allow an estimation of the spatial and seasonal phytoplankton biomass distribution in the surface mixed layer. The Samuels and Cox monthly global climatologies of the MLD (prepared from the NOAA database) were obtained from the National Center for Atmospheric Research website (http://dss.ucar.edu/catalogs). MLD data sets were available at a 1 1 resolution. The atmospheric mixing layer (AML) corresponding to the marine boundary layer was assumed to be constant and equal to 1000 m (Van Drooge et al., 2002). Even though this assumption may not be realistic in some regions, this assumption does not modify the conclusions of this study as discussed below. The present study is based on monthly values obtained from averaging the values corresponding to three consecutive years (1998, 1999, 2000) in order to obtain a climatologicallike data set of biogeophysical variables. 2.2. Latitudinal profiles of physical and biogeochemical variables Latitudinally averaged profiles of SST, chlorophyll a and MLD values for two representative months (January and July), are shown in Fig. 1. In contrast to the smooth temperature profile (Fig. 1a), the latitudinal profile of chlorophyll a concentrations (Fig. 1b) shows substantial variability, with pronounced and changing gradients over the whole profile, particularly at mid-high latitudes. Chlorophyll a (Fig. 1b) exhibits the patchiness usually encountered in biological variables, corresponding to a non-uniform spatial distribution of phytoplankton biomass (Valiela, 1995). Seasonality is higher than for temperature: in January there is more phytoplankton biomass in the southern oceans, whereas in July productivity increases in high latitudes of the northern hemisphere. The extend of the MLD (Fig. 1c) largely determines the nutrient supply within the photic zone and how deeply the phytoplankton is carried by vertical mixing of water (Valiela, 1995). The MLD is greater in zones with formation of deep water and poorer water column stratification, such as the North Atlantic ocean in winter (Apel, 1990). The latitudinal profile of the MLD has a stronger seasonality than temperature and chlorophyll a. However, along and around the equator, MLD values remain fairly constant all year round.
151
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
Due to the observed seasonality and spatial variability of the physical and biological data, temporal and spatial changes in the capacity of the surface oceans as a reservoir of PCBs can be expected. 2.3. Field measurements of atmospheric PCB concentrations Gas phase measurements of PCB concentrations have been reported for a north–south Atlantic transect from Germany to Chile (46N–46S) during the fall (October– December) of 1990. Details on the cruise route, sampling, and analytical procedures are given elsewhere (Schreitmu¨ller et al., 1994). These atmospheric concentrations have been assumed to be representative of those found in the world oceans. Concentrations at latitudes higher than those reported in this study have been extrapolated assuming they decrease steadily towards the poles. This simplification is supported by the well known temperature dependence of gas phase concentrations, and compilations of measurements that show a decrease with latitude (Axelman and Gustafsson, 2002). The Arctic ocean is not covered in this study.
3. Model description 3.1. Inventory of PCBs in the surface oceans
Fig. 1. Latitudinally averaged profiles of the biogeophysical data in the Atlantic ocean, in January and July (climatological months): (a) Temperature, (b) Chlorophyll a concentrations, (c) MLD.
The extent of the MLD may influence the vertical transport of PCBs, an issue that has not been studied in detail, so far. For example, a large MLD would cause PCBs to be in contact with the deep layers of the ocean. In contrast, a shallow MLD would cause concentrations in surface water to increase as a result of air–water exchange (Dachs et al., 1999).
The inventory (ng m2) of a given POP in the surface mixed layer of the oceanic water column is given by the sum of the concentration of the chemical in the dissolved (CW, ng m3) and particulate phases (CP, ng m3), multiplied by the mixed layer depth (MLD, m). Only the MLD of the water column is considered because it is the seawater mass under the direct influence of air–water exchange. In most cases, phytoplankton uptake may control the aquatic partitioning of POPs (Swackhamer and Skoglund, 1991) and phytoplankton accounts for most of the organic matter in the photic water column (Gasol et al., 1997). Furthermore, it has been shown that the vertical and spatial variability of particulate-phase POPs such as PCBs and PAHs follow the phytoplankton biomass variability (Dachs et al., 1997 a,b). In addition, it is considered that Cp is only relevant in the photic zone of the MLD (approximately 0–100 m). Therefore, the POP inventory in the MLD [ng m2] is given by: INVENTORY ¼ ðCW þ CP Þ MLD
ð1Þ
MLD 4 100 m INVENTORY ¼ CW MLD þ Cp 100 MLD > 100 m
152
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
where Cw (ng m3) and Cp (ng m3) are the POP concentrations of the dissolved and particulate (mainly phytoplankton) phases, respectively. 3.2. Ratio of maximum reservoir capacities (MRC) The atmosphere is the main transport vector and diffusive air–water exchange the main input route to the water column for PCBs and other POPs which atmospheric occurrence is mainly in the gas phase (Dachs et al., 1999; Van Drooge et al., 2002). Indeed, the dry particle deposition of aerosol associated PCBs is negligible as an atmosphere ocean transfer process of PCBs (Dachs et al., 2002; Jurado and Dachs, unpublished data). Therefore, the maximum reservoir capacity (MRC) of the surface ocean to hold PCBs will be given by the ratio of the inventory in the surface ocean to the amount of PCBs in the atmospheric mixed layer. Cw þ Cp MLD MRC ¼ MLD 4 100 m ð2Þ Cg AML MRC ¼
Cw MLD þ Cp 100 Cg AML
MLD > 100 m
where Cg (ng m3) is the POP gas phase concentration. Aerosol phase concentrations are assumed to be negligible in comparison to those in the gas phase and thus not contributing to the atmospheric inventory of PCBs. Estimated MRC values from Eq. (2) assume equilibrium between phases, consistent with a Level I fugacity model. This ratio allows an estimation of the total loading of a given PCB in the surface ocean water column, derived from an atmospheric concentration. It is important to stress that the inventory derived in this way will be an estimate of the maximum capacity of the surface oceans to hold PCBs. 3.2.1. Level I fugacity model The chemical is assumed to be at equilibrium between the gaseous, dissolved and particulate (mainly phytoplankton) phases. Specifically, PCBs in phytoplankton will move towards equilibrium with the dissolved phase; the latter will approach equilibrium with the gas phase. Air-water and water-phytoplankton exchange are viewed as reversible processes. Therefore, only the fugacitydriven processes of diffusive air–water exchange and phytoplankton uptake are considered. It is important to note that phytoplankton uptake is the first transfer step to food webs; the results presented here can therefore also be linked to consider the exposure of aquatic food webs to POPs. The biological pump (sinking of phytoplankton associated pollutants) and oceanic circulation are processes that work against equilibrium in the ocean (Murray, 1992), causing the real inventory of PCBs in surface
oceans to be presumably lower than that given by the MRC ratio. Vertical sinking of particle-associated pollutants is not considered in the level I predictions model because it is a non-fugacity driven process, but it will be considered in Section 5 below. Equilibrium occurs when the fugacity of each phase matches the rest of the fugacities (Mackay and Paterson, 1981; Mackay, 2001). Fugacity can be regarded as the ‘‘escaping tendency’’ of a chemical substance from a given phase and has units of pressure. At equilibrium: fg ¼ fw ¼ fp
ð3Þ
where fg (Pa) is the fugacity in the gas phase, fw (Pa), in water and fp (Pa), in the phytoplankton biomass. Fugacity is linearly related to concentration, by means of a fugacity ‘‘phase-specific’’ capacity constant: Z. This term quantifies the capacity of a given phase to hold the chemical, so that toxic substances tend to accumulate in phases where Z is high. It is thus possible to write a fugacity–concentration relationship in each phase: Cg ¼ fg Zg MW 109 Cw ¼ fw Zw MW 109 Cp ¼ fp Zp MW 10
ð4Þ
9
where MW is the chemical molecular weight (g mol1) and Zg, Zw and Zp are the ‘‘fugacity capacities’’ (mol Pa1 m3) in the gas, dissolved and phytoplankton phases, respectively, given by: Zg ¼
1 RT
ð5Þ
Zw ¼
1 H
ð6Þ
Zp ¼
r COC BCF H
ð7Þ
where R is the ideal gas constant (8.314 Pa m3 mol1 K1), T is the temperature (K), H is the Henry’s law constant (Pa m3 mol1), r is the ratio between the amount of organic matter and the amount of organic carbon [kgOM/ kgOC], COC is the concentration of organic carbon (kgOC m3), and BCF is the PCB bioconcentration factor in phytoplankton (m3/kg phytoplankton). The temperature dependence of the dimensionless Henry’s Law constant (H’) is described by the Gibbs– Helmholtz equation: ln H 0 ¼ -
DHH DSH þ RT R
ð8Þ
where dHH and dSH, (J mol1), are the enthalpy and entropy of the phase change from the dissolved phase to the gas phase and are independent of temperature. These are obtained from Bamford et al. (2000, 2002). Dissolved salts increase the value of H. In the case of PCBs, studied
153
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
here, this effect can be parameterized using a salting out constant of 0.3 (Schwarzenbach et al., 1993) by: 37 H 0sal ¼ H 0 exp 0:3 ð9Þ 58:5 Chlorophyll a concentrations may be used to estimate the concentration of organic carbon (COC) and hence the phytoplankton biomass (Valiela, 1995; Gasol et al., 1997). The ratio of chlorophyll to organic carbon is assumed to be 3103 (kgalgae chlorophyll/kgOC) (Jorgensen et al., 2000) and the value of r used is 2 kgOM/kgOC (Hedges et al., 2002). Phytoplankton is supposed to account for the majority of organic matter as discussed above. The BCF is the ratio between the dissolved and phytoplankton PCB concentrations at equilibrium. In order to obtain the BCF, it should be noted that phytoplankton uptake is modeled using a two compartment system: first there is a fast adsorption to the phytoplankton surface (BCFs), then a diffusion into the matrix via partitioning (BCFm). Hence, the overall BCF is expressed as the sum of the BCFm and BCFs (Skoglund et al., 1996; Dachs et al., 1999, 2000; Del Vento and Dachs, 2002). BCF ¼ BCFs þ BCFm
ð10Þ
BCFs and BCFm are related to compound physical-chemical properties, such as the octanol-water partition coefficient (KOW) at 298 K (Hawker and Connell, 1988). log BCFm ð298Þ ¼ 1:085 logKOW 3:770
ð11Þ
logKOW < 6:4 log BCFm ð298Þ ¼ 0:343 logKOW þ 0:913 logKOW 4 6:4 BCFs ð298Þ ¼ 233:61 logKOW 1084:05
ð12Þ
logKOW < 6:3 BCFs ð298Þ ¼ 398 6:3 4 logKOW < 7:0 BCFs ð298Þ ¼ 285:61 logKOW þ 2401:52 logKOW 5 7:0 Temperature affects bioaccumulation and uptake kinetics by microorganisms. The influence of temperature on the bioconcentration factor is given in the following expression: BCFðTÞ DHs 1 1 ¼ exp BCFð298Þ T 298 R
3.2.2. Air–water-phytoplankton equilibrium [(Cp+Cw)/Cg] A first approximation to the study of the variables affecting the MRC can be obtained from considering the ratio (Cp+Cw)/Cg which shows the tendency of the chemical to partition in aquatic phases in comparison to the gas phase. This enables the influence of temperature and phytoplankton biomass to be assessed, without taking into account the variability associated with MLD and AML. Using Eqs. (3) and (4), and assuming equilibrium conditions, the following expression is inferred: Cw þ Cp Zw þ Zp ¼ Cg Zg
ð14Þ
Replacing each fugacity capacity given in expressions (5), (6) and (7) gives: Cw þ Cp 1 þ r COC BCF ¼ H0 Cg
ð15Þ
3.2.3. Estimation of the ratio of maximum reservoir capacities (MRC) For the estimation of MRC, in addition to temperature and phytoplankton biomass, two new variables need to be considered: the mixed layer depth in the water column (MLD) and the atmospheric mixed layer depth (AML). As discussed above, MLD is obtained from a climatological data set, whilst AML is assumed to be constant and equal to 1000 m (Van Drooge et al., 2002). This assumption does not affect the conclusions of this study since it is mainly focused on the marine side rather than the atmospheric. Thus, assuming equilibrium in all phases and combining Eqs. (2) and (15), MRC can be expressed as a function of the concentration of organic matter, temperature, MLD and the physico-chemical properties of the PCBs: Zw þ Zp MLD MRC ¼ Zg AML ð1 þ r COC BCFÞ MLD ¼ H 0 ML MLD 4 100m Zw MLD þ Zp 100 MRC ¼ Zg AML MLD þ r COC BCF 100 ¼ H 0 AML MLD > 100m
ð16Þ
4. Results and discussion ð13Þ
where #Hs (KJ/mol) is the enthalpy of sorption, considered 35 KJ/mol for all the PCB compounds.
4.1. Spatial and seasonal variability of (Cw+Cp)/Cg For the calculations in this section, PCB 28 (3 chlorine atoms) and PCB 153 (6 chlorine atoms) have been
154
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
selected as indicative of PCBs with different physical– chemical properties. Fig. 2 shows the latitudinally averaged profiles of the maxima, minima and mean values of (Cw+Cp)/Cg for PCB 28 and 153 in the Atlantic ocean, in July. Examination of this ratio first, rather than studying MRC directly, facilitates an understanding of the latitudinal and seasonal behavior of MRC without considering the MLD, which sometimes masks the results due to its high variability in some regions. At first glance, Fig. 2 shows strong bimodal latitudinal trends, with greater values for this ratio at mid-high latitudes. A drop in the ratio is particularly pronounced in the subtropical regions. According to the cold condensation scenario, POPs volatilize from warm and temperate zones, undergo long range atmospheric
transport and subsequently recondense at colder, higher latitudes (Wania and Mackay 1996). Thus, this theory predicts lower (Cw+Cp)/Cg values at low latitudes and higher (Cw+Cp)/Cg values at high latitudes. Therefore, the observed variability in Fig. 2 is only partially consistent with the temperature-driven latitudinal profiles suggested by the global distillation hypothesis. Indeed, the latitudinal profiles in Fig. 2 show a large difference between maxima and minima, notably at high latitudes, and important gradients and variability throughout the profile. The largest peak is not located in the highest latitude, as would happen if the ratio was solely governed by temperature; rather it is located around 60 N. Furthermore, at a given latitude, there is a 10 fold variability in the (Cw+Cp)/Cg values, mainly as a con-
Fig. 2. Maps and latitudinally averaged profiles of mean, maximum and minimum of (Cw+Cp)/Cg in July for: (A) PCB 28, (B) PCB 153.
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
155
sequence of phytoplankton patchiness. This important spatial variability can not be predicted taking into account solely the influence of temperature. A closer examination of the averaged profile for PCB 28 (Fig. 2) shows variations between the maximum and minimum of a factor of 4–10, whereas for PCB 153 this is a factor of 100–1000. The increase of (Cw+Cp)/Cg in mid-high latitudes is greater for the more hydrophobic PCB because of the major role of phytoplankton uptake. The spatial, and specially the latitudinal variability of this ratio results from the dual variability in temperature and phytoplankton biomass. It is difficult to distinguish the effect of those two variables, because both increase at mid-high latitudes. An study of the latitudinal variability of temperature and phytoplankton biomass and their influence on the (Cw+Cp)/Cg ratio has been performed.
The seasonal trends at the mid-high latitude area follow a similar pattern to that of the chlorophyll a concentration, with a marked peak in spring (May–June) during the phytoplankton bloom. Lower values are predicted for the late summer, due to a decrease of phytoplankton biomass and higher temperatures. The area located in a subtropical region (15 N) is driven mostly by the effect of temperature, which has a dominant influence on H. Indeed, the profile is little altered during the bloom and shows a negative correlation with temperature. The minimum occurs in July–August and the maximum in December–January. In this area, phytoplankton biomass is very low and thus its seasonal variability is also of limited importance (see Figs. 2 and 3). Conversely, in the high productivity region, the (Cw+Cp)/Cg ratio is at a maximum during the spring phytoplankton bloom overwhelming the opposite influence of temperature.
4.1.1. Latitudinal variability of (Cw+Cp)/Cg that can be attributed to the variability of input parameters The influence of temperature and chlorophyll a on the latitudinal profile of (Cw+Cp)/Cg was investigated by varying one parameter at a time, whilst maintaining the other constant. Hence, to assess the variability attributed to temperature, the latitudinal profile of (Cw+Cp)/ Cg was estimated by using the average value of chlorophyll a (0.75 mg/m3) in the oceans (Dachs et al., 2002). Conversely, the variability attributed to phytoplankton biomass (i.e. chlorophyll concentration) was considered using a constant value of temperature of 286 K (Dachs et al., 2002). Results for PCB 28 and 153 are shown in Fig. 3a and b respectively. Chlorophyll a concentration produces a ‘spikiness’ in the latitudinal response, while temperature has a smoother effect latitudinally. An examination of individual plots suggests that for the more hydrophobic PCBs, the effect of biomass is stronger. On the other hand, temperature has a greater influence between 30 and 60 —in the zone where the gradient of T is greatest. Chlorophyll a concentration exerts its influence across the rest of the latitude range. Phytoplankton biomass is also responsible for the observed spatial patchiness. Thus, it is clear that temperature and compound physico-chemical properties are insufficient by themselves to predict the reservoir capacity of the surface oceans, especially for the more hydrophobic congeners.
4.2. Spatial and seasonal variability of the MRC
4.1.2. Seasonal variability of (Cw+Cp)/Cg The predicted seasonal trends of (Cw+Cp)/Cg are shown in Fig. 4a for PCB 153 in two contrasting areas of the north eastern Atlantic Ocean, one of high productivity (at 50 N), the other of low productivity (at 15 N). Average chlorophyll a concentrations and MLD in the two differentiate climatic regions are: 0.84 mg m3 and 86.7 m (50 N); 0.47 mg m3 and 22.1 m (15 N), respectively.
4.2.1. Seasonal trends of the MRC Fig. 4b exhibits the annual profile of the MRC in a high productivity area around 50 N and in a low productivity area around 15 N. In the low productivity area, in early spring the MLD decreases, leading to lower MRC values during late spring and summer. However, for the high productivity area this decrease occurs after the spring phytoplankton bloom. Therefore, for the high productivity zone, MRC has a similar pattern to the seasonal variation in the amount of biomass and MLD, while for the low productivity area, MRC is largely driven by the effect of temperature and MLD. 4.2.2. Spatial variability of the MRC Fig. 5 shows the spatially variability and the latitudinally averaged profiles of the MRC for PCB 28 and PCB 153. To illustrate the importance of seasonal variations of the MLD, the MRC calculated for January and July are compared. As observed for (Cw+Cp)/Cg (Fig. 2), a bimodal MRC trend is observed, with higher values at mid-high latitudes. The MRC peaks around 60 , due to higher values of phytoplankton biomass and MLD at this latitude. Fig. 5 shows that the factors driving the MLD variability have a strong influence on the MRC which has not considered in previous studies. The longitudinal variability of MRC appears to be of relatively minor importance than when just the (Cw+Cp)/Cg ratio is considered; however, it is still important. MRC patchiness due to phytoplankton biomass seems to be partially ‘‘damped’’ by the MLD when looking at the average latitudinal MRC values. MRC ratios are greater for PCB 153 than for PCB 28 (by about a factor of 10), in particular in mid-high latitudes. This is in agreement with the higher BCF values of PCB 153 than PCB 28.
156
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
Fig. 3. Latitudinally averaged profiles of the variability of (Cw+Cp)/Cg for PCB 28 and 153, in January and July: (a) Variability attributed to the variability of temperature, maintaining a constant value of concentration of chlorophyll a (0.75 mg chlorophyll/m3). (b) Variability attributed to the variability of chlorophyll a concentration, maintaining a constant value of temperature (286 K).
4.2.3. Influence of MLD variability on the latitudinal profiles of the MRC Using a constant value of MLD of 41 m (Dachs et al., 2002), the variability of MRC attributed to temperature and to chlorophyll a concentration is very similar to that observed earlier for (Cw+Cp)/Cg. Therefore, only the variability attributed to MLD, as shown in Fig. 6, is assessed in this section—the MLD exhibits strong seasonality with higher values during early winter and minimum values during summer stratification of the water column. Therefore, MLD influence on MRC is also seasonally dependent and values for January and July are shown. MLD spatial variability drives a strong increase by a factor of 7–8 of MRC values during winter. This dependence is important at mid-high latitudes while it is not important at the subtropical region. 4.2.4. Settling fluxes decrease the water column reservoir of PCBs The maximum potential inventory of PCBs in the oceanic mixed layer can be obtained by multiplying the MRC
by the gas-phase concentrations. In this study, the PCB gas-phase concentrations measured in a north–south transect have been used as a reference (Schreitmu¨ller et al., 1994). This maximum potential inventory, which is shown in Fig. 7, has been compared with that obtained from level III model that considers settling fluxes as a major removal process of PCBs from the water column. Indeed, recently, a level III model that assumes steady state conditions has been applied to the world’s oceans, to predict the processes controlling the oceanic sink of PCBs (Dachs et al., 2002). This model takes into account air–water exchange, phytoplankton uptake and settling of organic matter. Air–water fluxes (FAW, ng m2 d1) can be modeled in the traditional manner: FAW ¼ kAW
Cg Cw H0
ð17Þ
where kAW is the air–water mass transfer rate (m d1) that depends on temperature, wind speed and the phy-
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
157
Fig. 4. Seasonal variation for PCB 153 in two areas of 22 degree, one located in a high productivity region (around 50 N, 0.91 mg chlorophyll/m3), the other one in a low productivity region (around 15 N, 0.09 mg chlorphyll/m3): (a) (Cw+Cp)/Cg, (b) MRC.
sical–chemical properties of the pollutant. Fluxes of PCBs between water and plankton (FWP, ng m2 d1) are given by the following equation: kd FWP ¼ kWP Cw Cp ð18Þ ku where ku (m3 kg1 d1) and kd (d1) are the uptake and depuration rate constants, respectively. The mass transfer coefficient between water and plankton, kWP (m d1), depends on the mixed layer depth, the phytoplankton biomass, and the uptake constant (Dachs et al., 1999; Del Vento and Dachs, 2002). Vertical fluxes of particles and associated PCBs are parameterized by: Fsink ¼ ksink
kd Cp ku
ð19Þ
where kSink (m d1) is the mass transfer rate of sinking PCBs, related to the vertical flux of organic carbon. In this model, the sequestration of PCBs by the ocean is viewed as an air–water–plankton-deep water exchange process. On a long-term basis and due to compound persistence, the air–water–plankton-deep
ocean transport is assumed to be at steady state and therefore, FAW equals FWP and FSink. The system of Eqs. (17)–(19) is solved, making it possible to estimate the global variability of CW, CP, and thus of the surface ocean inventory of PCBs. In this case, equilibrium conditions are not assumed and settling of PCBs associated with sinking organic matter is considered as the major removal process. Fig. 7 also shows the predicted surface ocean inventory using this level III model for PCB 28 and PCB 153. Obviously, in both cases the inventory profile obtained by considering equilibrium conditions is higher than that obtained from dynamic (level III) conditions. This difference is very substantial for more chlorinated PCBs and in mid-high latitudes and upwelling areas. This observation provides evidence that settling of PCBs sorbed to organic matter is an important removal process from the surface oceans, especially for the more hydrophobic PCBs in high primary productivity areas, and is consistent with observations of lower concentrations in eutrophic aquatic ecosystems (Taylor et al., 1991; Skei et al., 2000) . Conversely, in the subtropical areas, due to low settling fluxes, the water
158
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
Fig. 5. Maps and latitudinally averaged profiles of MRC in January and July for: (A) PCB 28, (B) PCB 153.
column inventory is predicted to be close to equilibrium for all PCB congeners.
5. Relative importance of the surface ocean as a reservoir of PCBs Surface water inventories have been integrated for all the world oceans and for seven representative PCB congeners (PCBs 28, 52, 101, 118, 138, 153, 180) as predicted from the level III model. These inventories have been compared with those reported for soils at the global scale (Meijer et al., 2002) as shown in Fig. 8a.
The magnitude of these inventories is obviously dependent on the magnitude and distribution of CG used when applying the level III model [Eqs. (17)–(19)] and the uncertainty of these gas-phase concentrations affect the predictions shown in Fig. 8a. The concentrations and latitudinal distributions reported elsewhere (Schreitmu¨ller et al., 1994) are consistent with other measurements of gas-phase concentrations made over the ocean and in oceanic islands (Iwata et al., 1993; Lohmann et al., 2001; Van Drooge et al., 2002) and presumably the uncertainty associated to these concentrations and the predicted inventory in the surface ocean is lower than a factor of two.
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
159
Fig. 6. Latitudinally averaged profile of the variability of MRC attributed to the variability of the MLD, maintaining constant temperature and chlorophyll a concentration (286 K, 0.75 mg chlorophyll/ m3) for: (a) PCB 28, (b) PCB 153.
Fig. 7. Comparison of latitudinally averaged profiles of the global inventory in surface oceans ((Cw+Cp)*MLD) (ng/m2) as predicted when considering equilibrium and a level III model. (a) PCB28, (b) PCB 153.
The global PCB inventory in soils is greater than in the surface ocean for all congeners, but differences are more remarkable for the highly chlorinated PCBs. This suggests that the ocean is enriched in the lower chlorinated PCBs, which are not effectively removed by the settling of biogenic particulate matter. Conversely, the more hydrophobic congeners would become depleted in the water column due to efficient removal by settling particulate matter. Indeed, measured PCB profiles in soils are different from those predicted in surface oceans (Meijer et al., 2003; Dachs et al., 1997a,b). The inventories shown in Fig. 8 are for the surface ocean, i.e. above the MLD. However, the deeper ocean may contain a much greater inventory of PCBs, that remain to be quantified. In any case, oceanic deep water can be considered as a sink for PCBs since they will not move back to the atmosphere and cycle through the biogeosphere. The relative contribution of the individual PCBs in the predicted surface water inventories integrated for all the world oceans has been compared with that obtained
in other areas such as the North-Atlantic Ocean and the Mediterranean Sea (Dachs et al., 1997a; SchulzBull et al., 1998). For these field studies, the relative contribution of the measured PCBs in suspension (Cp) and in solution (Cw) has been considered. Results are shown in Fig. 8b. The PCB profiles derived here have similar trends to those reported for the Atlantic Ocean by Schulz-Bull et al. (1998). Some differences can presumably be attributed to variability of phytoplankton biomass in surface waters, causing variations in the partitioning and thus the congener specific PCB profile. Samples for the Mediterranean Sea (Dachs et al., 1997a) were dominated by the tri- and tetra-substituted species, which is consistent with a large and rapid sinking of particle associated pollutants in this area and their closer proximity to land sources. Samples from Schulz-Bull et al. (1998) showed very high concentrations in suspension compared to those in solution, as a result of the relatively high suspended matter concentrations in the sampling area (Northern Atlantic Ocean).
160
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
Fig. 8. (a) Comparison of the predicted inventories of PCBs in the surface ocean (using level III models) with that in background surface (0–5 cm) soils. (b) Predicted and measured relative contribution of PCB congeners in surface sea waters.
6. Concluding remarks The study of the biogeophysical variables that control the maximum capacity of the oceans to hold POPs such as PCBs shows that in high productivity areas, phytoplankton biomass and the extend of the MLD are the most important parameters. Conversely, in oligotrophic areas, temperature and the MLD drive the MRC values. This study has shown that it is important to assess the fate and occurrence of POPs in marine environment taking into account the biogeochemistry of POPs, spe-
cially their interactions with biota which leads to an important spatial and temporal variability. The role of the extend of the MLD on the fate and occurrence of POPs is an issue that needs to be assessed in detail. Furthermore, this study provides evidence that the surface ocean is enriched in the lower chlorinated PCBs, which are not effectively removed by the settling of biogenic particulate matter. This observation further support the hypothesis that long range atmospheric transport followed by air–water exchange and settling fluxes are key processes affecting the inventory and fate
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
of PCBs in the surface oceans. Furthermore, at the global scale, it can be hypothesized that continent-ocean interactions may play an important role of the global dynamics of PCBs and be responsible for the differences observed between soil and seawater profiles (Fig. 8). However, further research is needed to better assess transport routes and to understand the interactions between continents and oceans in terms of POP occurrence and transport.
Acknowledgements Dr. Rafel Simo´ is kindly acknowledged for fruitful discussions and as coordinator of the AMIGOS project. The authors would like to thank the SeaWiFS Project team and particularly W. Gregg from the Goddard Space Flight Center (NASA) for providing the satellite chlorophyll climatology and reviewers, who provided very valuable comments. This work was supported by the Spanish Ministry of Scince and Technology through project AMIGOS (REN2001-3462/CLI).
References Apel, J.R., 1990. Principles of Ocean Physics. Academic Press. Axelman, J., Gustafsson, O., 2002. Global sinks of PCBs: a critical assessment of the vapour-phase hydroxy radical sink emphasizing field diagnostic and model assumptions. Global Biogeochemical Cycles 16 (4), 58/1–58/13. Bamford, H.A., Poster, D.L., Baker, J.E., 2000. Henry’s Law constants of polychlorinated biphenyl congeners and their variation with temperature. J. Chem. Eng. Data 45, 1069–1074. Bamford, H.A., Poster, D.L., Huie, R.E., Baker, J.E., 2002. Using extrathermodynamic relationships to model the temperature dependence of Henry’s Law constants of 209 PCB congeners. Environmental Science Technology 36 (2), 4395–4402. Dachs, J., Bayona, J.M., Albaiges, J., 1997. Spatial distribution, vertical profiles and budget of organochlorine compounds in Western Mediterranean seawater. Marine Chemistry 57 (3–4), 313–324. Dachs, J., Bayona, J.M., Raoux, C., Albaige´s, J., 1997. Spatial, vertical distribution and budget of Polycyclic Aromatic Hydrocarbons in Western Mediterranean seawater. Environmental Science Technology 31 (3), 682–688. Dachs, J., Eisenreich, S.J., Baker, J.E., Ko, F.-C., Jeremiason, J.D., 1999. Coupling of phytoplankton uptake and air–water exchange of persistent organic pollutants. Environmental Science Technology 33 (20), 3653–3660. Dachs, J., Eisenreich, S.J., Hoff, R.M., 2000. Influence of eutrophication on air–water exchange, vertical fluxes, and phytoplankton concentrations of persistent organic pollutants. Environmental Science Technology 34 (6), 1095–1102. Dachs, J., Lohmann, R., Ockenden, W.A., Me´janelle, L., Eisenreich, S.J., Jones, K.C., 2002. Oceanic biogeochemical controls on global dynamics of persistent organic pollutants. Environmental Science Technology 36 (20), 4229–4237. Del Vento, S., Dachs, J., 2002. Prediction of uptake dynamics of persistent organic pollutants by bacteria and phytoplankton. Environmental Toxicology and Chemistry 21 (10), 2099–3017. Eisenberg, J.N.S., Bennett, D.H., Mckone, T.E., 1998. Chemical
161
dynamics of Persistent Organic Polluytants: a sensitivity analysis relating soil concentration levels to atmospheric emissions. Environmental Science Technology 32 (1), 115–123. Gasol, J., Del Giorgio, P., Duarte, C., 1997. Biomass distribution in marine planktonic communities. Limnol Oceanogr 42 (6), 1353– 1363. Hawker, D.W., Connell, D.W., 1988. Octanol-Water partition coefficients of polychlorinated biphenyl congeners. Environmental Science Technology 22 (4), 382–387. Hedges, J.I., Baldock, J.A., Gelinas, Y., Lee, C., Peterson, M.L., Wakeham, S.G., 2002. The biochemical and elemental compositions of marine plankton: a NMR perspective. Marine Chemistry 78 (1), 47–63. Iwata, I., Tanabe, S., Sakai, N., Tatsukawa, R., 1993. Distribution of persistent organochlorines in the oceanic air and surface seawater and the role of ocean on their global transport and fate. Environmental Science Technology 27, 1080–1098. Jorgensen, L.A., Jorgensen, S.E., Nielsen, S.N., 2000. ECOTOX: Ecological Modelling and Ecotoxicology. Elsevier, Amsterdam. Lohmann, R., Ockenden, W.A., Shears, J., Jones, K.C., 2001. Atmopheric distribution of polychlorinated dibenzo-p-dioxins, dibenzofurans (PCDD/Fs), and non-ortho biphenyls (PCBs) along a north– south transect. Environmental Science Technology 35 (20), 4046– 4053. Mackay, D., 2001. Multimedia Environmental Models. The Fugacity Approach. Lewis Publishers. Mackay, D., Paterson, S., 1981. Calculating fugacity. Environmental Science Technology 15 (9), 1006–1014. Meijer, S. N., Ockenden, W. A., Sweetman, A. J., Breivik, K., Grimalt, J. O., Jones, K. C.,2002. Global distribution and budget of PCBs and HCB in background surface soils: implications for sources and environmental processes. Environmental Science Technology (submitted). Murray, J.W., 1992. The Oceans. Global Biogeochemical Cycles. A. P. Limited. Scheringer, M., Wegmann, F., Fenner, K., Hungerbu¨hler, K., 2000. Investigation of the cold condensation of persistent organic pollutants with a global multimedia fate model. Environmental Science Technology 34, 1842–1850. Schreitmu¨ller, J., Vigneron, M., Bacher, R., Ballschmiter, K., 1994. Pattern analysis of polychlorinated biphenyls (PCB) in marine air of Atlantic ocean. Int. J. Environ. Anal. Chem. 57, 33–52. Schulz-Bull, D.E., Petrick, G., Bruhn, R., Duinker, J.C., 1998. Chlorobiphenyls (PCB) and PAHs in water masses of the northern North Atlantic. Marine Chemistry 61 (1–2), 101–114. Schwarzenbach, R.P., Gschwend, P.M., Imboden, D.M., 1993. Environmental Organic Chemistry. Wiley-Interscience. Skei, J., Larsson, P., Rosenberg, R., Jonsson, P., Olsson, M., Broman, D., 2000. Eutrophication and contaminants in aquatic ecosystems. Ambio 29 (4–5), 184–194. Skoglund, R.S., Stange, K., Swackhamer, D., 1996. A kinetics model for predicting the accumulation of PCBs in phytoplankton. Environmental Science Technology 30 (7), 2113–2120. Swackhamer, D.L., Skoglund, R.S., 1991. The role of phytoplankton in the partitioning of hydrophobic organic contaminants in water. Lewis publishers II. Taylor, W.D., Carey, J.H., Lean, D.R.S., McQueen, D.J., 1991. Organochlorine concentrations in the plankton of lakes in southern Ontario and their relationship to plankton biomass. Can. J. Fish. Aquat. Sci. 48, 1960–1966. Tolosa, I., Readman, J.W., Fowler, S.W., Villeneuve, J.P., Dachs, J., Bayona, J.M., Albaiges, J., 1997. PCBs in the western Mediterranean. Temporal trends and mass balance assessment. Deep-Sea Research II 44 (3–4), 907–928. Valiela, I., 1995. Marine Ecological Processes. Springer, New York. Van Drooge, B.L., Grimalt, J.O., Torres Garcia, C.J., Cuevas, E., 2002. Semivolatile organochlorine compounds in the free tropo-
162
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
sphere of the Northeastern Atlantic. Environmental Science Technology 36 (6), 1155–1161. Wania, F., Daly, G.L., 2002. Estimating the contribution of degradation in air and deposition to the deep sea to the global loss of PCBs. Atmospheric Environment 36 (36–37), 5581–5593. Wania, F., Mackay, D., 1996. Tracking the distribution of persistent
organic pollutants. Environmental Science Technology 30 (9), 390. A-396A. Wania, F., McLachlan, M.S., 2001. Estimating the influence of forests on the overall fate of semivolatile organic compounds using a multimedia fate model. Environmental Science Technology 35 (3), 582–590.
Latitudinal and seasonal capacity of the surface oceans as a reservoir of polychlorinated biphenyls Elena Juradoa, Rainer Lohmannb, Sandra Meijera,b, Kevin C. Jonesb, Jordi Dachsa,* a
Department of Environmental Chemistry, IIQAB-CSIC, Jordi Girona 18-26, Barcelona 08034, Catalunya, Spain b Environmental Science Department, Lancaster University, Lancaster LA1 4YQ, UK Received 14 July 2003; accepted 11 August 2003
‘‘Capsule’’: Model calculations estimate the latitudinal and seasonal storage capacity of the surface oceans for PCBs. Abstract The oceans play an important role as a global reservoir and ultimate sink of persistent organic pollutants (POPs) such as polychlorinated biphenyls congeners (PCBs). However, the physical and biogeochemical variables that affect the oceanic capacity to retain PCBs show an important spatial and temporal variability which have not been studied in detail, so far. The objective of this paper is to assess the seasonal and spatial variability of the ocean’s maximum capacity to act as a reservoir of atmospherically transported and deposited PCBs. A level I fugacity model is used which incorporates the environmental variables of temperature, phytoplankton biomass, and mixed layer depth, as determined from remote sensing and from climatological datasets. It is shown that temperature, phytoplankton biomass and mixed layer depth influence the potential PCB reservoir of the oceans, being phytoplankton biomass specially important in the oceanic productive regions. The ocean’s maximum capacities to hold PCBs are estimated. They are compared to a budget of PCBs in the surface oceans derived using a level III model that assumes steady state and which incorporates water column settling fluxes as a loss process. Results suggest that settling fluxes will keep the surface oceanic reservoir of PCBs well below its maximum capacity, especially for the more hydrophobic compounds. The strong seasonal and latitudinal variability of the surface ocean’s storage capacity needs further research, because it plays an important role in the global biogeochemical cycles controlling the ultimate sink of PCBs. Because this modeling exercise incorporates variations in downward fluxes driven by phytoplankton and the extent of the water column mixing, it predicts more complex latitudinal variations in PCBs concentrations than those previously suggested. # 2003 Elsevier Ltd. All rights reserved. Keywords: POP; PCB; Global dynamics; Marine pollution; Fugacity model
1. Introduction The oceans play an important role in controlling the environmental transport, fate and sinks of persistent organic pollutants (POPs) at regional and global scales (Wania and Mackay, 1996; Dachs et al., 2002). Even though POP concentrations in the open ocean are lower than those observed in coastal areas (Iwata et al., 1993; Dachs et al., 1997a; Schulz-Bull et al., 1998), the large oceanic volumes imply that they may represent an important inventory of POPs. This has been confirmed in budgets performed for some marine regions, such as the Western Mediterranean, and may
* Corresponding author. Tel.: +34-93-4006118. E-mail address: [email protected] (J. Dachs). 0269-7491/$ - see front matter # 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.envpol.2003.08.039
be true for the global oceans (Dachs et al., 1997a; Tolosa et al., 1997). Wania and Mackay (1996) suggested that temperature plays a significant role in determining the transport and sinking of POPs in the oceans at the global scale, through the processes of cold condensation, global distillation and latitudinal fractionation. This would result in an inverted concentration gradient of POPs, with low concentrations in the tropics and high concentrations in the polar regions. On the other hand, less volatile compounds are deposited close to their sources, while more volatile compounds travel further toward the poles, where they deposit preferentially due to lower temperatures. Therefore it has been suggested that temperature could largely explain the oceanic concentrations of POPs due to its influence on air–water exchange and other partitioning processes (Wania and Mackay, 1996).
150
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
However, due to the hydrophobic character of POPs, the spatial and temporal variability in the distribution of water column organic matter has an important influence on the oceanic inventory of POPs. Recently, it has been shown that phytoplankton uptake and settling of particulate matter is a driver of the oceanic sink of POPs such as polychlorinated biphenyls (PCBs) and dibenzo-p-dioxins and-furans (PCDD/Fs) (Dachs et al., 1999, 2000, 2002; Skei et al., 2000; Wania and Daly, 2002) and that the deep ocean is an important global sink for the more hydrophobic POPs (Wania and Daly, 2002). Besides the role that organic matter plays on the settling of POPs, no assessment exists of the potential influence of the spatial variability of organic matter and other biogeophysical variables on the capacity of the surface oceans as a reservoir of POPs. The physical and biogeochemical variables with potential influence on the oceanic inventory of POPs, such as phytoplankton biomass and mixed layer depth, show an important spatial and seasonal variability which presumably affects the cycling and long term fate of POPs. Soil and vegetation organic matter, such as forested surfaces, has traditionally been considered in models that study the global distribution of POPs in terrestrial environments (Eisenberg et al., 1998; Scheringer et al., 2000; Wania and McLachlan, 2001). These studies have shown that terrestrial organic matter has an important influence on the dynamics and inventory of POPs. Therefore, it seems logical that high productivity oceanic areas may exert an important influence on the global distribution of POPs. Specifically, this work will be focused in POPs, such as PCBs, that are found preferentially in the gas-phase (Van Drooge et al., 2002). The objectives of this work are to study the influence of environmental variables such as temperature, biomass and mixed layer depth on the reservoir capacity of PCBs by the suface oceans. The role of settling fluxes driven by enhanced phytoplankton productivity is assessed as a key factor affecting the global oceanic inventory of PCBs. The results are compared with those predicted by the global distillation theory under equilibrium conditions. Finally, the predicted inventory of PCBs in the surface ocean is compared with that reported in soils.
2. Data sources and field sampling 2.1. Physical and biogeochemical data The global distribution of sea surface temperature (SST), chlorophyll a concentrations, and the mixed layer depth (MLD) were assessed as potential variables affecting the reservoir capacity of oceans on a global scale. Data on their global distribution and variability were obtained by remote sensing from satellites and climatological datasets. SSTs were obtained from the Along Track Scanning Radiometer (ATSR) installed in
the European Space Agency ERS-2 satellite (ATSR project web page http://www.atsr.rl.ac.uk/). SST images consist of monthly averaged data with a resolution of a half degree and an accuracy of +/0.3 K. Chlorophyll a concentrations were estimated from fluorescence signals obtained from the Sea-viewing Wide Field-of-view Sensor (SeaWIFS) mission funded by NASA’s Mission To Planet Earth (MTPE) Program (http://seawifs.gsfc.nasa.gov/SEAWIFS.html). Data are averaged monthly at 1 1 resolution and an estimated error of 15%. Chlorophyll a concentrations allow an estimation of the spatial and seasonal phytoplankton biomass distribution in the surface mixed layer. The Samuels and Cox monthly global climatologies of the MLD (prepared from the NOAA database) were obtained from the National Center for Atmospheric Research website (http://dss.ucar.edu/catalogs). MLD data sets were available at a 1 1 resolution. The atmospheric mixing layer (AML) corresponding to the marine boundary layer was assumed to be constant and equal to 1000 m (Van Drooge et al., 2002). Even though this assumption may not be realistic in some regions, this assumption does not modify the conclusions of this study as discussed below. The present study is based on monthly values obtained from averaging the values corresponding to three consecutive years (1998, 1999, 2000) in order to obtain a climatologicallike data set of biogeophysical variables. 2.2. Latitudinal profiles of physical and biogeochemical variables Latitudinally averaged profiles of SST, chlorophyll a and MLD values for two representative months (January and July), are shown in Fig. 1. In contrast to the smooth temperature profile (Fig. 1a), the latitudinal profile of chlorophyll a concentrations (Fig. 1b) shows substantial variability, with pronounced and changing gradients over the whole profile, particularly at mid-high latitudes. Chlorophyll a (Fig. 1b) exhibits the patchiness usually encountered in biological variables, corresponding to a non-uniform spatial distribution of phytoplankton biomass (Valiela, 1995). Seasonality is higher than for temperature: in January there is more phytoplankton biomass in the southern oceans, whereas in July productivity increases in high latitudes of the northern hemisphere. The extend of the MLD (Fig. 1c) largely determines the nutrient supply within the photic zone and how deeply the phytoplankton is carried by vertical mixing of water (Valiela, 1995). The MLD is greater in zones with formation of deep water and poorer water column stratification, such as the North Atlantic ocean in winter (Apel, 1990). The latitudinal profile of the MLD has a stronger seasonality than temperature and chlorophyll a. However, along and around the equator, MLD values remain fairly constant all year round.
151
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
Due to the observed seasonality and spatial variability of the physical and biological data, temporal and spatial changes in the capacity of the surface oceans as a reservoir of PCBs can be expected. 2.3. Field measurements of atmospheric PCB concentrations Gas phase measurements of PCB concentrations have been reported for a north–south Atlantic transect from Germany to Chile (46N–46S) during the fall (October– December) of 1990. Details on the cruise route, sampling, and analytical procedures are given elsewhere (Schreitmu¨ller et al., 1994). These atmospheric concentrations have been assumed to be representative of those found in the world oceans. Concentrations at latitudes higher than those reported in this study have been extrapolated assuming they decrease steadily towards the poles. This simplification is supported by the well known temperature dependence of gas phase concentrations, and compilations of measurements that show a decrease with latitude (Axelman and Gustafsson, 2002). The Arctic ocean is not covered in this study.
3. Model description 3.1. Inventory of PCBs in the surface oceans
Fig. 1. Latitudinally averaged profiles of the biogeophysical data in the Atlantic ocean, in January and July (climatological months): (a) Temperature, (b) Chlorophyll a concentrations, (c) MLD.
The extent of the MLD may influence the vertical transport of PCBs, an issue that has not been studied in detail, so far. For example, a large MLD would cause PCBs to be in contact with the deep layers of the ocean. In contrast, a shallow MLD would cause concentrations in surface water to increase as a result of air–water exchange (Dachs et al., 1999).
The inventory (ng m2) of a given POP in the surface mixed layer of the oceanic water column is given by the sum of the concentration of the chemical in the dissolved (CW, ng m3) and particulate phases (CP, ng m3), multiplied by the mixed layer depth (MLD, m). Only the MLD of the water column is considered because it is the seawater mass under the direct influence of air–water exchange. In most cases, phytoplankton uptake may control the aquatic partitioning of POPs (Swackhamer and Skoglund, 1991) and phytoplankton accounts for most of the organic matter in the photic water column (Gasol et al., 1997). Furthermore, it has been shown that the vertical and spatial variability of particulate-phase POPs such as PCBs and PAHs follow the phytoplankton biomass variability (Dachs et al., 1997 a,b). In addition, it is considered that Cp is only relevant in the photic zone of the MLD (approximately 0–100 m). Therefore, the POP inventory in the MLD [ng m2] is given by: INVENTORY ¼ ðCW þ CP Þ MLD
ð1Þ
MLD 4 100 m INVENTORY ¼ CW MLD þ Cp 100 MLD > 100 m
152
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
where Cw (ng m3) and Cp (ng m3) are the POP concentrations of the dissolved and particulate (mainly phytoplankton) phases, respectively. 3.2. Ratio of maximum reservoir capacities (MRC) The atmosphere is the main transport vector and diffusive air–water exchange the main input route to the water column for PCBs and other POPs which atmospheric occurrence is mainly in the gas phase (Dachs et al., 1999; Van Drooge et al., 2002). Indeed, the dry particle deposition of aerosol associated PCBs is negligible as an atmosphere ocean transfer process of PCBs (Dachs et al., 2002; Jurado and Dachs, unpublished data). Therefore, the maximum reservoir capacity (MRC) of the surface ocean to hold PCBs will be given by the ratio of the inventory in the surface ocean to the amount of PCBs in the atmospheric mixed layer. Cw þ Cp MLD MRC ¼ MLD 4 100 m ð2Þ Cg AML MRC ¼
Cw MLD þ Cp 100 Cg AML
MLD > 100 m
where Cg (ng m3) is the POP gas phase concentration. Aerosol phase concentrations are assumed to be negligible in comparison to those in the gas phase and thus not contributing to the atmospheric inventory of PCBs. Estimated MRC values from Eq. (2) assume equilibrium between phases, consistent with a Level I fugacity model. This ratio allows an estimation of the total loading of a given PCB in the surface ocean water column, derived from an atmospheric concentration. It is important to stress that the inventory derived in this way will be an estimate of the maximum capacity of the surface oceans to hold PCBs. 3.2.1. Level I fugacity model The chemical is assumed to be at equilibrium between the gaseous, dissolved and particulate (mainly phytoplankton) phases. Specifically, PCBs in phytoplankton will move towards equilibrium with the dissolved phase; the latter will approach equilibrium with the gas phase. Air-water and water-phytoplankton exchange are viewed as reversible processes. Therefore, only the fugacitydriven processes of diffusive air–water exchange and phytoplankton uptake are considered. It is important to note that phytoplankton uptake is the first transfer step to food webs; the results presented here can therefore also be linked to consider the exposure of aquatic food webs to POPs. The biological pump (sinking of phytoplankton associated pollutants) and oceanic circulation are processes that work against equilibrium in the ocean (Murray, 1992), causing the real inventory of PCBs in surface
oceans to be presumably lower than that given by the MRC ratio. Vertical sinking of particle-associated pollutants is not considered in the level I predictions model because it is a non-fugacity driven process, but it will be considered in Section 5 below. Equilibrium occurs when the fugacity of each phase matches the rest of the fugacities (Mackay and Paterson, 1981; Mackay, 2001). Fugacity can be regarded as the ‘‘escaping tendency’’ of a chemical substance from a given phase and has units of pressure. At equilibrium: fg ¼ fw ¼ fp
ð3Þ
where fg (Pa) is the fugacity in the gas phase, fw (Pa), in water and fp (Pa), in the phytoplankton biomass. Fugacity is linearly related to concentration, by means of a fugacity ‘‘phase-specific’’ capacity constant: Z. This term quantifies the capacity of a given phase to hold the chemical, so that toxic substances tend to accumulate in phases where Z is high. It is thus possible to write a fugacity–concentration relationship in each phase: Cg ¼ fg Zg MW 109 Cw ¼ fw Zw MW 109 Cp ¼ fp Zp MW 10
ð4Þ
9
where MW is the chemical molecular weight (g mol1) and Zg, Zw and Zp are the ‘‘fugacity capacities’’ (mol Pa1 m3) in the gas, dissolved and phytoplankton phases, respectively, given by: Zg ¼
1 RT
ð5Þ
Zw ¼
1 H
ð6Þ
Zp ¼
r COC BCF H
ð7Þ
where R is the ideal gas constant (8.314 Pa m3 mol1 K1), T is the temperature (K), H is the Henry’s law constant (Pa m3 mol1), r is the ratio between the amount of organic matter and the amount of organic carbon [kgOM/ kgOC], COC is the concentration of organic carbon (kgOC m3), and BCF is the PCB bioconcentration factor in phytoplankton (m3/kg phytoplankton). The temperature dependence of the dimensionless Henry’s Law constant (H’) is described by the Gibbs– Helmholtz equation: ln H 0 ¼ -
DHH DSH þ RT R
ð8Þ
where dHH and dSH, (J mol1), are the enthalpy and entropy of the phase change from the dissolved phase to the gas phase and are independent of temperature. These are obtained from Bamford et al. (2000, 2002). Dissolved salts increase the value of H. In the case of PCBs, studied
153
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
here, this effect can be parameterized using a salting out constant of 0.3 (Schwarzenbach et al., 1993) by: 37 H 0sal ¼ H 0 exp 0:3 ð9Þ 58:5 Chlorophyll a concentrations may be used to estimate the concentration of organic carbon (COC) and hence the phytoplankton biomass (Valiela, 1995; Gasol et al., 1997). The ratio of chlorophyll to organic carbon is assumed to be 3103 (kgalgae chlorophyll/kgOC) (Jorgensen et al., 2000) and the value of r used is 2 kgOM/kgOC (Hedges et al., 2002). Phytoplankton is supposed to account for the majority of organic matter as discussed above. The BCF is the ratio between the dissolved and phytoplankton PCB concentrations at equilibrium. In order to obtain the BCF, it should be noted that phytoplankton uptake is modeled using a two compartment system: first there is a fast adsorption to the phytoplankton surface (BCFs), then a diffusion into the matrix via partitioning (BCFm). Hence, the overall BCF is expressed as the sum of the BCFm and BCFs (Skoglund et al., 1996; Dachs et al., 1999, 2000; Del Vento and Dachs, 2002). BCF ¼ BCFs þ BCFm
ð10Þ
BCFs and BCFm are related to compound physical-chemical properties, such as the octanol-water partition coefficient (KOW) at 298 K (Hawker and Connell, 1988). log BCFm ð298Þ ¼ 1:085 logKOW 3:770
ð11Þ
logKOW < 6:4 log BCFm ð298Þ ¼ 0:343 logKOW þ 0:913 logKOW 4 6:4 BCFs ð298Þ ¼ 233:61 logKOW 1084:05
ð12Þ
logKOW < 6:3 BCFs ð298Þ ¼ 398 6:3 4 logKOW < 7:0 BCFs ð298Þ ¼ 285:61 logKOW þ 2401:52 logKOW 5 7:0 Temperature affects bioaccumulation and uptake kinetics by microorganisms. The influence of temperature on the bioconcentration factor is given in the following expression: BCFðTÞ DHs 1 1 ¼ exp BCFð298Þ T 298 R
3.2.2. Air–water-phytoplankton equilibrium [(Cp+Cw)/Cg] A first approximation to the study of the variables affecting the MRC can be obtained from considering the ratio (Cp+Cw)/Cg which shows the tendency of the chemical to partition in aquatic phases in comparison to the gas phase. This enables the influence of temperature and phytoplankton biomass to be assessed, without taking into account the variability associated with MLD and AML. Using Eqs. (3) and (4), and assuming equilibrium conditions, the following expression is inferred: Cw þ Cp Zw þ Zp ¼ Cg Zg
ð14Þ
Replacing each fugacity capacity given in expressions (5), (6) and (7) gives: Cw þ Cp 1 þ r COC BCF ¼ H0 Cg
ð15Þ
3.2.3. Estimation of the ratio of maximum reservoir capacities (MRC) For the estimation of MRC, in addition to temperature and phytoplankton biomass, two new variables need to be considered: the mixed layer depth in the water column (MLD) and the atmospheric mixed layer depth (AML). As discussed above, MLD is obtained from a climatological data set, whilst AML is assumed to be constant and equal to 1000 m (Van Drooge et al., 2002). This assumption does not affect the conclusions of this study since it is mainly focused on the marine side rather than the atmospheric. Thus, assuming equilibrium in all phases and combining Eqs. (2) and (15), MRC can be expressed as a function of the concentration of organic matter, temperature, MLD and the physico-chemical properties of the PCBs: Zw þ Zp MLD MRC ¼ Zg AML ð1 þ r COC BCFÞ MLD ¼ H 0 ML MLD 4 100m Zw MLD þ Zp 100 MRC ¼ Zg AML MLD þ r COC BCF 100 ¼ H 0 AML MLD > 100m
ð16Þ
4. Results and discussion ð13Þ
where #Hs (KJ/mol) is the enthalpy of sorption, considered 35 KJ/mol for all the PCB compounds.
4.1. Spatial and seasonal variability of (Cw+Cp)/Cg For the calculations in this section, PCB 28 (3 chlorine atoms) and PCB 153 (6 chlorine atoms) have been
154
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
selected as indicative of PCBs with different physical– chemical properties. Fig. 2 shows the latitudinally averaged profiles of the maxima, minima and mean values of (Cw+Cp)/Cg for PCB 28 and 153 in the Atlantic ocean, in July. Examination of this ratio first, rather than studying MRC directly, facilitates an understanding of the latitudinal and seasonal behavior of MRC without considering the MLD, which sometimes masks the results due to its high variability in some regions. At first glance, Fig. 2 shows strong bimodal latitudinal trends, with greater values for this ratio at mid-high latitudes. A drop in the ratio is particularly pronounced in the subtropical regions. According to the cold condensation scenario, POPs volatilize from warm and temperate zones, undergo long range atmospheric
transport and subsequently recondense at colder, higher latitudes (Wania and Mackay 1996). Thus, this theory predicts lower (Cw+Cp)/Cg values at low latitudes and higher (Cw+Cp)/Cg values at high latitudes. Therefore, the observed variability in Fig. 2 is only partially consistent with the temperature-driven latitudinal profiles suggested by the global distillation hypothesis. Indeed, the latitudinal profiles in Fig. 2 show a large difference between maxima and minima, notably at high latitudes, and important gradients and variability throughout the profile. The largest peak is not located in the highest latitude, as would happen if the ratio was solely governed by temperature; rather it is located around 60 N. Furthermore, at a given latitude, there is a 10 fold variability in the (Cw+Cp)/Cg values, mainly as a con-
Fig. 2. Maps and latitudinally averaged profiles of mean, maximum and minimum of (Cw+Cp)/Cg in July for: (A) PCB 28, (B) PCB 153.
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
155
sequence of phytoplankton patchiness. This important spatial variability can not be predicted taking into account solely the influence of temperature. A closer examination of the averaged profile for PCB 28 (Fig. 2) shows variations between the maximum and minimum of a factor of 4–10, whereas for PCB 153 this is a factor of 100–1000. The increase of (Cw+Cp)/Cg in mid-high latitudes is greater for the more hydrophobic PCB because of the major role of phytoplankton uptake. The spatial, and specially the latitudinal variability of this ratio results from the dual variability in temperature and phytoplankton biomass. It is difficult to distinguish the effect of those two variables, because both increase at mid-high latitudes. An study of the latitudinal variability of temperature and phytoplankton biomass and their influence on the (Cw+Cp)/Cg ratio has been performed.
The seasonal trends at the mid-high latitude area follow a similar pattern to that of the chlorophyll a concentration, with a marked peak in spring (May–June) during the phytoplankton bloom. Lower values are predicted for the late summer, due to a decrease of phytoplankton biomass and higher temperatures. The area located in a subtropical region (15 N) is driven mostly by the effect of temperature, which has a dominant influence on H. Indeed, the profile is little altered during the bloom and shows a negative correlation with temperature. The minimum occurs in July–August and the maximum in December–January. In this area, phytoplankton biomass is very low and thus its seasonal variability is also of limited importance (see Figs. 2 and 3). Conversely, in the high productivity region, the (Cw+Cp)/Cg ratio is at a maximum during the spring phytoplankton bloom overwhelming the opposite influence of temperature.
4.1.1. Latitudinal variability of (Cw+Cp)/Cg that can be attributed to the variability of input parameters The influence of temperature and chlorophyll a on the latitudinal profile of (Cw+Cp)/Cg was investigated by varying one parameter at a time, whilst maintaining the other constant. Hence, to assess the variability attributed to temperature, the latitudinal profile of (Cw+Cp)/ Cg was estimated by using the average value of chlorophyll a (0.75 mg/m3) in the oceans (Dachs et al., 2002). Conversely, the variability attributed to phytoplankton biomass (i.e. chlorophyll concentration) was considered using a constant value of temperature of 286 K (Dachs et al., 2002). Results for PCB 28 and 153 are shown in Fig. 3a and b respectively. Chlorophyll a concentration produces a ‘spikiness’ in the latitudinal response, while temperature has a smoother effect latitudinally. An examination of individual plots suggests that for the more hydrophobic PCBs, the effect of biomass is stronger. On the other hand, temperature has a greater influence between 30 and 60 —in the zone where the gradient of T is greatest. Chlorophyll a concentration exerts its influence across the rest of the latitude range. Phytoplankton biomass is also responsible for the observed spatial patchiness. Thus, it is clear that temperature and compound physico-chemical properties are insufficient by themselves to predict the reservoir capacity of the surface oceans, especially for the more hydrophobic congeners.
4.2. Spatial and seasonal variability of the MRC
4.1.2. Seasonal variability of (Cw+Cp)/Cg The predicted seasonal trends of (Cw+Cp)/Cg are shown in Fig. 4a for PCB 153 in two contrasting areas of the north eastern Atlantic Ocean, one of high productivity (at 50 N), the other of low productivity (at 15 N). Average chlorophyll a concentrations and MLD in the two differentiate climatic regions are: 0.84 mg m3 and 86.7 m (50 N); 0.47 mg m3 and 22.1 m (15 N), respectively.
4.2.1. Seasonal trends of the MRC Fig. 4b exhibits the annual profile of the MRC in a high productivity area around 50 N and in a low productivity area around 15 N. In the low productivity area, in early spring the MLD decreases, leading to lower MRC values during late spring and summer. However, for the high productivity area this decrease occurs after the spring phytoplankton bloom. Therefore, for the high productivity zone, MRC has a similar pattern to the seasonal variation in the amount of biomass and MLD, while for the low productivity area, MRC is largely driven by the effect of temperature and MLD. 4.2.2. Spatial variability of the MRC Fig. 5 shows the spatially variability and the latitudinally averaged profiles of the MRC for PCB 28 and PCB 153. To illustrate the importance of seasonal variations of the MLD, the MRC calculated for January and July are compared. As observed for (Cw+Cp)/Cg (Fig. 2), a bimodal MRC trend is observed, with higher values at mid-high latitudes. The MRC peaks around 60 , due to higher values of phytoplankton biomass and MLD at this latitude. Fig. 5 shows that the factors driving the MLD variability have a strong influence on the MRC which has not considered in previous studies. The longitudinal variability of MRC appears to be of relatively minor importance than when just the (Cw+Cp)/Cg ratio is considered; however, it is still important. MRC patchiness due to phytoplankton biomass seems to be partially ‘‘damped’’ by the MLD when looking at the average latitudinal MRC values. MRC ratios are greater for PCB 153 than for PCB 28 (by about a factor of 10), in particular in mid-high latitudes. This is in agreement with the higher BCF values of PCB 153 than PCB 28.
156
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
Fig. 3. Latitudinally averaged profiles of the variability of (Cw+Cp)/Cg for PCB 28 and 153, in January and July: (a) Variability attributed to the variability of temperature, maintaining a constant value of concentration of chlorophyll a (0.75 mg chlorophyll/m3). (b) Variability attributed to the variability of chlorophyll a concentration, maintaining a constant value of temperature (286 K).
4.2.3. Influence of MLD variability on the latitudinal profiles of the MRC Using a constant value of MLD of 41 m (Dachs et al., 2002), the variability of MRC attributed to temperature and to chlorophyll a concentration is very similar to that observed earlier for (Cw+Cp)/Cg. Therefore, only the variability attributed to MLD, as shown in Fig. 6, is assessed in this section—the MLD exhibits strong seasonality with higher values during early winter and minimum values during summer stratification of the water column. Therefore, MLD influence on MRC is also seasonally dependent and values for January and July are shown. MLD spatial variability drives a strong increase by a factor of 7–8 of MRC values during winter. This dependence is important at mid-high latitudes while it is not important at the subtropical region. 4.2.4. Settling fluxes decrease the water column reservoir of PCBs The maximum potential inventory of PCBs in the oceanic mixed layer can be obtained by multiplying the MRC
by the gas-phase concentrations. In this study, the PCB gas-phase concentrations measured in a north–south transect have been used as a reference (Schreitmu¨ller et al., 1994). This maximum potential inventory, which is shown in Fig. 7, has been compared with that obtained from level III model that considers settling fluxes as a major removal process of PCBs from the water column. Indeed, recently, a level III model that assumes steady state conditions has been applied to the world’s oceans, to predict the processes controlling the oceanic sink of PCBs (Dachs et al., 2002). This model takes into account air–water exchange, phytoplankton uptake and settling of organic matter. Air–water fluxes (FAW, ng m2 d1) can be modeled in the traditional manner: FAW ¼ kAW
Cg Cw H0
ð17Þ
where kAW is the air–water mass transfer rate (m d1) that depends on temperature, wind speed and the phy-
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
157
Fig. 4. Seasonal variation for PCB 153 in two areas of 22 degree, one located in a high productivity region (around 50 N, 0.91 mg chlorophyll/m3), the other one in a low productivity region (around 15 N, 0.09 mg chlorphyll/m3): (a) (Cw+Cp)/Cg, (b) MRC.
sical–chemical properties of the pollutant. Fluxes of PCBs between water and plankton (FWP, ng m2 d1) are given by the following equation: kd FWP ¼ kWP Cw Cp ð18Þ ku where ku (m3 kg1 d1) and kd (d1) are the uptake and depuration rate constants, respectively. The mass transfer coefficient between water and plankton, kWP (m d1), depends on the mixed layer depth, the phytoplankton biomass, and the uptake constant (Dachs et al., 1999; Del Vento and Dachs, 2002). Vertical fluxes of particles and associated PCBs are parameterized by: Fsink ¼ ksink
kd Cp ku
ð19Þ
where kSink (m d1) is the mass transfer rate of sinking PCBs, related to the vertical flux of organic carbon. In this model, the sequestration of PCBs by the ocean is viewed as an air–water–plankton-deep water exchange process. On a long-term basis and due to compound persistence, the air–water–plankton-deep
ocean transport is assumed to be at steady state and therefore, FAW equals FWP and FSink. The system of Eqs. (17)–(19) is solved, making it possible to estimate the global variability of CW, CP, and thus of the surface ocean inventory of PCBs. In this case, equilibrium conditions are not assumed and settling of PCBs associated with sinking organic matter is considered as the major removal process. Fig. 7 also shows the predicted surface ocean inventory using this level III model for PCB 28 and PCB 153. Obviously, in both cases the inventory profile obtained by considering equilibrium conditions is higher than that obtained from dynamic (level III) conditions. This difference is very substantial for more chlorinated PCBs and in mid-high latitudes and upwelling areas. This observation provides evidence that settling of PCBs sorbed to organic matter is an important removal process from the surface oceans, especially for the more hydrophobic PCBs in high primary productivity areas, and is consistent with observations of lower concentrations in eutrophic aquatic ecosystems (Taylor et al., 1991; Skei et al., 2000) . Conversely, in the subtropical areas, due to low settling fluxes, the water
158
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
Fig. 5. Maps and latitudinally averaged profiles of MRC in January and July for: (A) PCB 28, (B) PCB 153.
column inventory is predicted to be close to equilibrium for all PCB congeners.
5. Relative importance of the surface ocean as a reservoir of PCBs Surface water inventories have been integrated for all the world oceans and for seven representative PCB congeners (PCBs 28, 52, 101, 118, 138, 153, 180) as predicted from the level III model. These inventories have been compared with those reported for soils at the global scale (Meijer et al., 2002) as shown in Fig. 8a.
The magnitude of these inventories is obviously dependent on the magnitude and distribution of CG used when applying the level III model [Eqs. (17)–(19)] and the uncertainty of these gas-phase concentrations affect the predictions shown in Fig. 8a. The concentrations and latitudinal distributions reported elsewhere (Schreitmu¨ller et al., 1994) are consistent with other measurements of gas-phase concentrations made over the ocean and in oceanic islands (Iwata et al., 1993; Lohmann et al., 2001; Van Drooge et al., 2002) and presumably the uncertainty associated to these concentrations and the predicted inventory in the surface ocean is lower than a factor of two.
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
159
Fig. 6. Latitudinally averaged profile of the variability of MRC attributed to the variability of the MLD, maintaining constant temperature and chlorophyll a concentration (286 K, 0.75 mg chlorophyll/ m3) for: (a) PCB 28, (b) PCB 153.
Fig. 7. Comparison of latitudinally averaged profiles of the global inventory in surface oceans ((Cw+Cp)*MLD) (ng/m2) as predicted when considering equilibrium and a level III model. (a) PCB28, (b) PCB 153.
The global PCB inventory in soils is greater than in the surface ocean for all congeners, but differences are more remarkable for the highly chlorinated PCBs. This suggests that the ocean is enriched in the lower chlorinated PCBs, which are not effectively removed by the settling of biogenic particulate matter. Conversely, the more hydrophobic congeners would become depleted in the water column due to efficient removal by settling particulate matter. Indeed, measured PCB profiles in soils are different from those predicted in surface oceans (Meijer et al., 2003; Dachs et al., 1997a,b). The inventories shown in Fig. 8 are for the surface ocean, i.e. above the MLD. However, the deeper ocean may contain a much greater inventory of PCBs, that remain to be quantified. In any case, oceanic deep water can be considered as a sink for PCBs since they will not move back to the atmosphere and cycle through the biogeosphere. The relative contribution of the individual PCBs in the predicted surface water inventories integrated for all the world oceans has been compared with that obtained
in other areas such as the North-Atlantic Ocean and the Mediterranean Sea (Dachs et al., 1997a; SchulzBull et al., 1998). For these field studies, the relative contribution of the measured PCBs in suspension (Cp) and in solution (Cw) has been considered. Results are shown in Fig. 8b. The PCB profiles derived here have similar trends to those reported for the Atlantic Ocean by Schulz-Bull et al. (1998). Some differences can presumably be attributed to variability of phytoplankton biomass in surface waters, causing variations in the partitioning and thus the congener specific PCB profile. Samples for the Mediterranean Sea (Dachs et al., 1997a) were dominated by the tri- and tetra-substituted species, which is consistent with a large and rapid sinking of particle associated pollutants in this area and their closer proximity to land sources. Samples from Schulz-Bull et al. (1998) showed very high concentrations in suspension compared to those in solution, as a result of the relatively high suspended matter concentrations in the sampling area (Northern Atlantic Ocean).
160
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
Fig. 8. (a) Comparison of the predicted inventories of PCBs in the surface ocean (using level III models) with that in background surface (0–5 cm) soils. (b) Predicted and measured relative contribution of PCB congeners in surface sea waters.
6. Concluding remarks The study of the biogeophysical variables that control the maximum capacity of the oceans to hold POPs such as PCBs shows that in high productivity areas, phytoplankton biomass and the extend of the MLD are the most important parameters. Conversely, in oligotrophic areas, temperature and the MLD drive the MRC values. This study has shown that it is important to assess the fate and occurrence of POPs in marine environment taking into account the biogeochemistry of POPs, spe-
cially their interactions with biota which leads to an important spatial and temporal variability. The role of the extend of the MLD on the fate and occurrence of POPs is an issue that needs to be assessed in detail. Furthermore, this study provides evidence that the surface ocean is enriched in the lower chlorinated PCBs, which are not effectively removed by the settling of biogenic particulate matter. This observation further support the hypothesis that long range atmospheric transport followed by air–water exchange and settling fluxes are key processes affecting the inventory and fate
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
of PCBs in the surface oceans. Furthermore, at the global scale, it can be hypothesized that continent-ocean interactions may play an important role of the global dynamics of PCBs and be responsible for the differences observed between soil and seawater profiles (Fig. 8). However, further research is needed to better assess transport routes and to understand the interactions between continents and oceans in terms of POP occurrence and transport.
Acknowledgements Dr. Rafel Simo´ is kindly acknowledged for fruitful discussions and as coordinator of the AMIGOS project. The authors would like to thank the SeaWiFS Project team and particularly W. Gregg from the Goddard Space Flight Center (NASA) for providing the satellite chlorophyll climatology and reviewers, who provided very valuable comments. This work was supported by the Spanish Ministry of Scince and Technology through project AMIGOS (REN2001-3462/CLI).
References Apel, J.R., 1990. Principles of Ocean Physics. Academic Press. Axelman, J., Gustafsson, O., 2002. Global sinks of PCBs: a critical assessment of the vapour-phase hydroxy radical sink emphasizing field diagnostic and model assumptions. Global Biogeochemical Cycles 16 (4), 58/1–58/13. Bamford, H.A., Poster, D.L., Baker, J.E., 2000. Henry’s Law constants of polychlorinated biphenyl congeners and their variation with temperature. J. Chem. Eng. Data 45, 1069–1074. Bamford, H.A., Poster, D.L., Huie, R.E., Baker, J.E., 2002. Using extrathermodynamic relationships to model the temperature dependence of Henry’s Law constants of 209 PCB congeners. Environmental Science Technology 36 (2), 4395–4402. Dachs, J., Bayona, J.M., Albaiges, J., 1997. Spatial distribution, vertical profiles and budget of organochlorine compounds in Western Mediterranean seawater. Marine Chemistry 57 (3–4), 313–324. Dachs, J., Bayona, J.M., Raoux, C., Albaige´s, J., 1997. Spatial, vertical distribution and budget of Polycyclic Aromatic Hydrocarbons in Western Mediterranean seawater. Environmental Science Technology 31 (3), 682–688. Dachs, J., Eisenreich, S.J., Baker, J.E., Ko, F.-C., Jeremiason, J.D., 1999. Coupling of phytoplankton uptake and air–water exchange of persistent organic pollutants. Environmental Science Technology 33 (20), 3653–3660. Dachs, J., Eisenreich, S.J., Hoff, R.M., 2000. Influence of eutrophication on air–water exchange, vertical fluxes, and phytoplankton concentrations of persistent organic pollutants. Environmental Science Technology 34 (6), 1095–1102. Dachs, J., Lohmann, R., Ockenden, W.A., Me´janelle, L., Eisenreich, S.J., Jones, K.C., 2002. Oceanic biogeochemical controls on global dynamics of persistent organic pollutants. Environmental Science Technology 36 (20), 4229–4237. Del Vento, S., Dachs, J., 2002. Prediction of uptake dynamics of persistent organic pollutants by bacteria and phytoplankton. Environmental Toxicology and Chemistry 21 (10), 2099–3017. Eisenberg, J.N.S., Bennett, D.H., Mckone, T.E., 1998. Chemical
161
dynamics of Persistent Organic Polluytants: a sensitivity analysis relating soil concentration levels to atmospheric emissions. Environmental Science Technology 32 (1), 115–123. Gasol, J., Del Giorgio, P., Duarte, C., 1997. Biomass distribution in marine planktonic communities. Limnol Oceanogr 42 (6), 1353– 1363. Hawker, D.W., Connell, D.W., 1988. Octanol-Water partition coefficients of polychlorinated biphenyl congeners. Environmental Science Technology 22 (4), 382–387. Hedges, J.I., Baldock, J.A., Gelinas, Y., Lee, C., Peterson, M.L., Wakeham, S.G., 2002. The biochemical and elemental compositions of marine plankton: a NMR perspective. Marine Chemistry 78 (1), 47–63. Iwata, I., Tanabe, S., Sakai, N., Tatsukawa, R., 1993. Distribution of persistent organochlorines in the oceanic air and surface seawater and the role of ocean on their global transport and fate. Environmental Science Technology 27, 1080–1098. Jorgensen, L.A., Jorgensen, S.E., Nielsen, S.N., 2000. ECOTOX: Ecological Modelling and Ecotoxicology. Elsevier, Amsterdam. Lohmann, R., Ockenden, W.A., Shears, J., Jones, K.C., 2001. Atmopheric distribution of polychlorinated dibenzo-p-dioxins, dibenzofurans (PCDD/Fs), and non-ortho biphenyls (PCBs) along a north– south transect. Environmental Science Technology 35 (20), 4046– 4053. Mackay, D., 2001. Multimedia Environmental Models. The Fugacity Approach. Lewis Publishers. Mackay, D., Paterson, S., 1981. Calculating fugacity. Environmental Science Technology 15 (9), 1006–1014. Meijer, S. N., Ockenden, W. A., Sweetman, A. J., Breivik, K., Grimalt, J. O., Jones, K. C.,2002. Global distribution and budget of PCBs and HCB in background surface soils: implications for sources and environmental processes. Environmental Science Technology (submitted). Murray, J.W., 1992. The Oceans. Global Biogeochemical Cycles. A. P. Limited. Scheringer, M., Wegmann, F., Fenner, K., Hungerbu¨hler, K., 2000. Investigation of the cold condensation of persistent organic pollutants with a global multimedia fate model. Environmental Science Technology 34, 1842–1850. Schreitmu¨ller, J., Vigneron, M., Bacher, R., Ballschmiter, K., 1994. Pattern analysis of polychlorinated biphenyls (PCB) in marine air of Atlantic ocean. Int. J. Environ. Anal. Chem. 57, 33–52. Schulz-Bull, D.E., Petrick, G., Bruhn, R., Duinker, J.C., 1998. Chlorobiphenyls (PCB) and PAHs in water masses of the northern North Atlantic. Marine Chemistry 61 (1–2), 101–114. Schwarzenbach, R.P., Gschwend, P.M., Imboden, D.M., 1993. Environmental Organic Chemistry. Wiley-Interscience. Skei, J., Larsson, P., Rosenberg, R., Jonsson, P., Olsson, M., Broman, D., 2000. Eutrophication and contaminants in aquatic ecosystems. Ambio 29 (4–5), 184–194. Skoglund, R.S., Stange, K., Swackhamer, D., 1996. A kinetics model for predicting the accumulation of PCBs in phytoplankton. Environmental Science Technology 30 (7), 2113–2120. Swackhamer, D.L., Skoglund, R.S., 1991. The role of phytoplankton in the partitioning of hydrophobic organic contaminants in water. Lewis publishers II. Taylor, W.D., Carey, J.H., Lean, D.R.S., McQueen, D.J., 1991. Organochlorine concentrations in the plankton of lakes in southern Ontario and their relationship to plankton biomass. Can. J. Fish. Aquat. Sci. 48, 1960–1966. Tolosa, I., Readman, J.W., Fowler, S.W., Villeneuve, J.P., Dachs, J., Bayona, J.M., Albaiges, J., 1997. PCBs in the western Mediterranean. Temporal trends and mass balance assessment. Deep-Sea Research II 44 (3–4), 907–928. Valiela, I., 1995. Marine Ecological Processes. Springer, New York. Van Drooge, B.L., Grimalt, J.O., Torres Garcia, C.J., Cuevas, E., 2002. Semivolatile organochlorine compounds in the free tropo-
162
E. Jurado et al. / Environmental Pollution 128 (2004) 149–162
sphere of the Northeastern Atlantic. Environmental Science Technology 36 (6), 1155–1161. Wania, F., Daly, G.L., 2002. Estimating the contribution of degradation in air and deposition to the deep sea to the global loss of PCBs. Atmospheric Environment 36 (36–37), 5581–5593. Wania, F., Mackay, D., 1996. Tracking the distribution of persistent
organic pollutants. Environmental Science Technology 30 (9), 390. A-396A. Wania, F., McLachlan, M.S., 2001. Estimating the influence of forests on the overall fate of semivolatile organic compounds using a multimedia fate model. Environmental Science Technology 35 (3), 582–590.